Dataset automatically created during the evaluation run of model huihui-ai/Qwen2.5-32B-Instruct-abliterated.
The dataset is composed of 136 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing… See the full description on the dataset page:
https://huggingface.co/datasets/OALL/details_huihui-ai__Qwen2.5-32B-Instruct-abliterated.
